Flow resistance in open channels, that is, the friction that the channel boundary exerts on the flow, is a fundamental hydraulic parameter defining many aspects of the behavior of a river channel (Kalathil & Chandra, 2019;Powell, 2014;Shobe et al., 2021). The roughness of the channel boundary relative to the depth of the flow is one of the primary influences on flow resistance (Cheng, 2017;Ferguson, 2007;Nitsche et al., 2012;Powell, 2014), and river channels can be broadly divided into two categories based on boundary roughness.The first category are those channels where the flow is significantly deeper than the scale of the boundary roughness, for example, deep relative to the characteristic grain size or bedform height. In these flows, the fluid momentum is dissipated in a roughness boundary layer that is thin relative to the full flow depth. The small scale of the roughness boundary layer relative to the depth of the flow ensures that the spatial variability in flow resistance is low, allowing the flow to reasonably be treated as steady and uniform. Overall, the effect of the boundary layer on the flow is well captured by a simple adjustment to the log law of the wall depth-velocity profile based on a roughness length-scale. Although the details of how any particular boundary configuration gives rise to a particular roughness length-scale are still poorly understood (Brereton et al., 2021;Chung et al., 2021), a length-scale of ∼3D 84 is commonly used for natural channels (Bathurst, 1985;Lamb et al., 2017a). The D 84 is the grain diameter larger than 84% of the grains on a bed, and is commonly used to characterize the grain sizes in a river channel. This is in contrast to the second category where the boundary roughness length-scale is as large or larger than the flow depth due to roughness elements such as immobile boulders, often found in steep mountain environments. The flow resistance for these channels is high and generally underpredicted by models developed for relatively smooth boundaries (Kalathil & Chandra, 2019;Powell, 2014;Shobe et al., 2021). Because the roughness boundary layer occupies much or all of the flow depth, the poor understanding of flow within this layer becomes especially problematic. In addition, the flow resistance tends to be spatially variable, varying from one channel cross-section to another, and the flow is unlikely to be steady or uniform. Despite significant research on the topic, there is not a widely accepted theory of flow resistance for channels with large roughness, for example, boulder-mantled channels (